<p>Autonomous underwater vehicles are fundamental technologies for exploring underwater environments, with applications in climatic monitoring, defense systems, structure integrity inspection, etc. However, the localization of underwater vehicles is still a complex task, mainly due to the difficulty of using conventional communication and sensory perception, commonly in autonomous navigation. Thus, this paper presents a survey on a novel way of underwater robots’ localization based on multi-domain images. In this sense, an overview of multi-domain image fusion methods applied for ground localization using aerial images is provided. It can inspire new underwater localization techniques that adopt aerial images and methods for knowledge extraction from aerial images of coastal regions. Furthermore, new approaches for underwater localization based on aerial images and image-enhanced techniques for acoustic underwater images are an important source of information for improving underwater localization methods.</p>